Approaches to Intelligent Systems


Event Details
Tuesday, October 8, 2013
Talk:
4:00 p.m., Avery 115

Reception:
3:30 p.m., Avery 348

Jens G. Pohl, Ph.D.

Professor Emeritus, California Polytechnic State University

Abstract

The presentation will examine the three prevalent approaches to Artificial Intelligence (AI), namely symbolic reasoning systems (top down approach), connectionist systems (bottom up approach), and emergent systems based on the principles of the subsumption theory. Distinguished by their top-down and bottom-up mechanisms all three approaches have strengths and weaknesses. While the logical reasoning approach is precise and well supported by mathematical theories and procedures, it is constrained by a largely predefined representational model. Connectionist systems, on the other hand, are able to recognize patterns even if these patterns are only similar and not identical to the patterns that they have been trained to recognize, but they have no understanding of the meaning of those patterns.  The subsumption approach  appears  to  overcome  many  of  the weaknesses of the other two approaches in theory, but there is concern that it may not scale to more complex real world applications.

 There are weaknesses that all three AI approaches share, namely inability to deal with exceptions, lack of mechanisms for analogous comparisons, and very primitive conceptualization capabilities at best. It is noted that the human agent performs decidedly better in these areas. The presentation will conclude with the proposition that only a hybrid approach holds sufficient promise to meet the full expectations of intelligent software systems. It is further suggested that this hybrid approach should in most cases include the contributions of the human agent as an integral component of the intelligent system.

 While the author will present a somewhat critical review of these three approaches to intelligent software systems, it is not at all intended to suggest that the application of these methods either individually or in combination is not useful in business domains. Even with their limitations they lead to business intelligence capabilities that are orders of magnitude more powerful than rote data processing techniques

Speaker Bio

Professor Jens Pohl (emeritus) retired from the California Polytechnic State University (Cal Poly, San Luis Obispo) in 2010, where he held the positions of Professor of Architecture, Executive Director of the Collaborative Agent Design Research Center (CADRC),  Director of the  Knowledge Management Laboratory (KML Center) at Scott AFB, IL, and  Graduate  Studies Coordinator, in the College of Architecture and Environmental Design. During the period 2002 to 2010 he concurrently held the positions of Chief Technology Officer and Chairman of the Board of Directors, CDM Technologies, Inc. (www.cdmtechnology.com), a California Corporation and sustaining sponsor of the CADRC Center at Cal Poly. Following the acquisition of CDM by Boeing in December 2010, Dr  Pohl became Vice President (Senior Technology Advisor) and Senior Director, Adaptive Systems ofTapestry Solutions (a Boeing Company) (www.tapestrysolutions.com).